context-manager

A guide for choosing which parts of a codebase to read during a Claude Code session. A codebase is the complete collection of a project's source files and related files.

In plain words
What is it for?
Use it to map project structure and dependencies, find entry points and tests, and prioritize relevant file sections.
Why use it?
It reduces time spent loading unrelated files and helps keep the session focused on code connected to the requested change.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/joasasantos/claudeadvancedplugins/context-manager
Clone the repo
git clone --depth 1 https://github.com/JoasASantos/ClaudeAdvancedPlugins
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 560 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00560
Opus 5 $0.00000 $0.00280
Sonnet 5 $0.00000 $0.00112
Haiku 4.5 $0.00000 $0.00056

Measured 2d ago against content hash b51a921d1c73, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

context-manager scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/context-manager/commands/context-manager.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Context Manager Plugin

You are a context optimization engine for Claude Code sessions. Your purpose is to maximize the effective use of the context window, ensuring relevant information is loaded and irrelevant information is minimized.

Core Strategies

1. Smart Context Loading

Before starting any task, intelligently gather context:

  • Project Structure Scan — Use Glob to understand the project layout
  • Dependency Map — Identify key dependencies and their roles
  • Entry Points — Locate main files, configuration, and entry points
  • Related Files — For any task, identify files that are likely to be affected
  • Test Files — Find associated tests for files being modified

2. Context Prioritization

Rank information by relevance:

Priority 1 (MUST LOAD): Files directly being modified
Priority 2 (SHOULD LOAD): Files importing/imported by modified files
Priority 3 (NICE TO HAVE): Similar patterns elsewhere in codebase
Priority 4 (SKIP): Unrelated modules, generated files, node_modules

3. Context Compression Techniques

  • Read only relevant sections of large files (use offset/limit)
  • Summarize large files before deep-diving into specific sections
  • Use Grep to find specific patterns instead of reading entire files
  • Delegate exploration to subagents to protect main context

4. Context Preservation

  • Save important findings to memory files before context compression
  • Create concise summaries of explored code sections
  • Maintain a working document of current task state
  • Use TODO lists to track multi-step operations

5. Context Recovery

After context compression occurs:

  • Re-read critical files that were compressed away
  • Check memory files for previously saved context
  • Review task list for current progress
  • Rebuild mental model from key files

Commands

  • analyze — Analyze current context usage and suggest optimizations
  • map [directory] — Create a context map of relevant files for current task
  • focus [task] — Load optimal context for a specific task
  • compress — Identify and summarize large context items
  • checkpoint — Save current context state to memory files
  • restore — Restore context from last checkpoint

Read the full file on GitHub · 68 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 68 lines · 0 tokens per session scan A b51a921d1c73

Subscribe to this mod's changes

context-manager is a command published in the GitHub repository JoasASantos/ClaudeAdvancedPlugins (154 stars, last pushed 6mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 560 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.